Success story · Banking · 2002
Reducing defaults without slowing sales
How a vague risk problem became a measurable chain of causes – and a German solution became a Europe-wide standard.
3.1 → 3.5
Process sigma of the first payment rate
- ≈ 3,000loan contracts due per month in scope
- DMAICfull Six Sigma cycle
According to the Green Belt project documentation, 2002
01 / 04 · Situation
The first instalment never arrives
For a share of newly disbursed loans, the first instalment never reached the loan account. There were plenty of assumptions and no reliable figures.
loan contracts due per month in scope
02 / 04 · Straight talk
Measure first, then discuss
Three independent reviewers cross-checked the measurement system. A blanket “risk problem” became five clearly separate causes, each with its own figure.
agreement between reviewers
03 / 04 · Execution
Implement before the report is finished
Four of the five measures went into the credit policy while the analysis was still running, together with a four-eyes check of bank details and a monthly exception list.
measures in force while the analysis was still running
04 / 04 · Result
Fewer defaults while new business grew
Measured over nine months. The insights were then built into GE Money Bank’s credit risk models across Europe.
Process sigma of the first payment rate · defaults on the first instalment ≈ −⅓
According to the Green Belt project documentation “First Payment Default”, 2002. Schematic illustration; a dot is not an individual case.
Starting point
For a share of newly disbursed loans, the first instalment never reached the loan account. For the bank, this was not a side issue but a problem with three costs: customers received reminders for something that was not their fault. Sales and back office reworked cases that should never have arisen. And each of these cases tied up loan loss provisions that hit the bottom line directly.
The scale was not academic: each individual default on the first instalment tied up around 3,200 euros in loan loss provisions on average, and over 4,000 euros in the intermediary business.
Above all, one thing was unclear: what was causing it? Assumptions ranged from data entry errors and overly generous credit decisions to customers who simply did not want to pay. There were no reliable figures – and therefore no basis for a decision that would have stood up in front of the board.
Mandate & role
Markus Slotta joined Service Bank, which later became part of GE Money Bank, in August 1999 and stayed until April 2005. He started in customer advisory, took over as head of the Bochum branch in February 2002 and, most recently, as Task Force Manager in Retail Sales, was responsible for the branch network in the Rhine-Main-Neckar region – initially 18, eventually up to 36 locations. This role in my career history →
In parallel, he led the Six Sigma project “First Payment Default” and was involved in the due diligence for the merger of Service Bank with Allbank. He was certified as a Green Belt for the FPD project – both are explicitly recorded in his employment reference.
Approach
Measure first, then discuss
Before a single measure was discussed, it had to be established that the figures were sound. The data came from the data warehouse; the measurement system itself was cross-checked through a Gage R&R analysis with three independent reviewers – with agreement of 98 and 99 per cent respectively. Only then did the discussion move beyond opinion.
Separate the causes, do not lump them together
A blanket “risk problem” became five clearly distinguishable causes – errors in capturing bank details, unverified secondary income, decisions at the margin of creditworthiness, age effects and differences between new and existing customers. Each cause was given its own figure. This made it visible that the biggest levers were not where the organisation had assumed them to be.
The same rigour when someone else’s figures are on the table
Alongside his line responsibility, Markus Slotta was involved in the due diligence for the merger of Service Bank with Allbank: valuation of the loan portfolios and review of the underlying collateral. The thinking is the same as in one’s own process – nothing is taken on trust that cannot be recalculated. The difference lies in the stakes: here, the valuation was the basis for a merger decision. And the work did not end with the review report, but only with the integration measures that followed.
Implement before the report is finished
Four of the five measures derived were not presented at the end of the project, but adopted into the credit policy as immediate measures while the analysis phase was still under way. In addition, there was a four-eyes check of bank details, documented in the file checklist, and a monthly exception list that gave the branches direct access to the affected accounts. The end result was not a slide deck but a signed handover sheet to the process owner. The insights from the project were subsequently incorporated into GE Money Bank’s credit risk models across Europe.
Result
over a measurement period of nine months – while new business was growing at the same time
The project formed the basis of Markus Slotta’s GE certification as a Six Sigma Green Belt (March 2003).
You do not reduce risk with stricter rules, but by understanding where errors actually arise – and fixing the process exactly there.
Your next step
Facing a transformation that needs more than slides?
A 30-minute first conversation. Straight talk, no sales routine.
Book a first conversation View success stories